.Net Engineer to Manage SaaS Project

Type: Full-time, senior/lead individual contributor with technical ownership (can grow into an engineering lead) What you’d own: The reliability, correctness, and roadmap of a live production platform that loan officers use every day to reach borrowers by SMS, email, and phone. Real money and real customer relationships ride on it working perfectly. The mission BeapBeap runs automated, multi-day communication campaigns (text, email, voice) for mortgage loan officers, plus a browser softphone, proposal generation, and lead intake from paid aggregators. Today it works, but it has grown fast and carries the kinds of problems that only surface at scale: a campaign silently stopping when a vendor rate-limits one message, a loan officer whose calls do not ring their cell because a provisioning step quietly failed, paid leads bought and never contacted, and “delivered” receipts that never reached the handset. Your job is to make this platform boringly reliable, observable, and safe to change, and to keep shipping features without breaking what already works. What you will actually do • Own the campaign orchestration engine. It runs on Azure Durable Functions. You need to deeply understand orchestration replay determinism (why changing orchestrator code under running workflows corrupts them), side-by-side versioning, and safe deploys that never strand in-flight runs. Recover stranded runs without re-sending messages people already got. • Own the telephony and messaging stack. We integrate a CPaaS vendor (SipShield/Telucom) for SMS/MMS, A2P 10DLC campaigns, carrier call-forwarding, ringless voicemail, WebRTC softphone, and call bridging. You will manage that integration end to end, diagnose carrier-level delivery failures, and be our technical point of contact with the vendor. • Make delivery provable, not assumed. Vendor “success” signals lie. You will build the telemetry and reconciliation so we can prove a text reached a handset, a call actually rang, and a number is correctly provisioned, instead of trusting an optimistic status. • Harden the deploy pipeline. Gated deploys, staging-slot swaps, canary checks, automatic rollback, and guards that measure against what is actually running in production, not against a diff. Deploys should be safe enough to do daily. • Fix root causes, not symptoms. When something breaks, trace it to the seed, fix that and every sibling call site, and add the standing check that would have caught it. No band-aids, no “add a warning” in place of a fix. • Data integrity. The system has multiple stores that can disagree (Cosmos containers, derived fields, vendor records). You will enforce single-writer invariants and build reconcilers that surface drift before it hurts a customer. • Observability and on-call. Stand up real monitoring (Application Insights / KQL, harm-signal alerting) so we see problems before customers report them, and own incident response when they happen. • Partner with the business. Work directly with the CEO and the loan-officer customers, translate vague reports (“my forward is broken”) into rooted diagnoses, and prioritize by customer and revenue impact. Must-have skills • C# / .NET (8+), ASP.NET Core production experience, several years, on a real revenue system. • Azure, hands-on: App Service and deployment slots, Azure Durable Functions (orchestrations, activities, replay model), Cosmos DB (partition keys, query cost, consistency), Azure CLI / az rest / ARM. Durable Functions replay understanding is close to non-negotiable, it is the highest-risk part of the system. • CPaaS / telephony integration: SMS/MMS delivery, A2P 10DLC registration and campaign caps, delivery receipts (DLRs), carrier call-forwarding, and ideally WebRTC and PSTN call bridging. You do not need to have built a carrier, but you must be comfortable reasoning about “the vendor said delivered but the phone never rang.” • Deployment and reliability engineering: CI/CD, blue-green / slot swaps, feature flags, safe rollback, and writing deploy guards that fail closed. • Observability: structured logging, metrics, distributed tracing, log-query languages (KQL or equivalent), building alerts that fire on the signal you would actually act on. • Debugging and root-cause discipline: the temperament to prove a claim with a query and its output rather than assume, to distrust a green dashboard, and to reproduce before fixing. • Testing rigor: unit and integration tests, plus the judgment to know when a passing test proves nothing (e.g. a test that asserts a shape but never exercises the real path). Strong plus • Mortgage, lending, or fintech domain experience (loan officers, borrowers, 1003 / loan applications, proposals, rate/APR math). • Compliance-aware messaging: TCPA, quiet hours, DNC/opt-out handling, A2P trust tiers. • Experience owning a third-party vendor relationship and pushing it to fix platform-level gaps. • Email deliverability (SPF/DKIM/DMARC, ESP integration). • Prior experience being the senior engineer on a small team or a solo owner of a production system. The kind of person who succeeds here • Takes end-to-end ownership: “it is not done until it is verified working in production for a real user.” • Communicates plainly with a non-engineer founder, gives a recommendation instead of a menu, and flags risk honestly. • Calm under a live incident, methodical, and allergic to guessing. • Comfortable moving fast on a system with real customers, and disciplined about not breaking it while doing so.